Discomfort from sinusoidal oscillation in the pitch and fore-and-aft ...

More documents

Recommendations

Info

464 ARTICLE IN PRESS The four equivalent comfort contours shown in Fig. 2, correspond to conditions in which there was the same discomfort at all frequencies—hence the root-sums-of-squares (i.e. r.s.s.) summation of weighted components for each of these four conditions should be the same if the standardised weighting method is appropriate—at least at frequencies greater than 0.5 Hz where it is intended to be used. Fig. 6 shows the various components to which subjects were exposed in the four conditions, weighted as suggested in BS6841 [10] (assuming the weightings are extrapolated to frequencies less than 0.5 Hz, see Fig. 7), together with the root-sums-of-squares of these weighted components so as to produce an ‘overall ride value’. The prediction method in BS 6841 [10] allows for the assessment of discomfort caused by complex motions containing both rotational and translational components using weightings that assume the rotational components are measured in rad s 2 and the translational components are measured in m s 2 . In principal, after frequency weighting according to the weightings and multiplying factors in the standard, components in any axis (translational or rotational) can be compared and combined (using root-sums-of-squares summation) such that the ‘overall ride value’ is appropriately influenced by each component according to its importance. It may be seen in Fig. 6 that for fore-and-aft oscillation without a backrest, the equivalent comfort contour (i.e. the r.s.s. of the weighted fore-and-aft seat and fore-and-aft floor accelerations) corresponds to a similar overall ride value at all frequencies from 0.5 to 1.6 Hz, as expected. The main contributor to the overall ride value is the fore-and-aft acceleration at the seat, with acceleration at the feet making a negligible contribution using the standardised method. However, the r.s.s. of the weighted contours appears to have underestimated the subject’s sensitivity as the contour level is somewhat below the level of the reference motion (0.2 Hz, 0.45 m s 2 rms in the fore-and-aft axis). Ride value (m.s -2 r.m.s.) 1 0.1 0.01 0.001 0.1 0.01 I.H. Wyllie, M.J. Griffin / Journal of Sound and Vibration 324 (2009) 453–467 seat pitch, We feet fore-and-aft, W b feet tangential, W b feet fore-and-aft, W b r.s.s. summation seat fore-&-aft, W d r.s.s. summation seat fore-&-aft, W d seat pitch, W e feet fore-and-aft, W b feet tangential, W b feet fore-and-aft, W b r.s.s. summation back tangential, W c seat fore-&-aft, W d back fore-&-aft, W c r.s.s. summation seat fore-&-aft, W d back fore-&-aft, W c 0.001 0.1 1 0.1 Frequency (Hz) 1 10 Fig. 6. Frequency-weighted accelerations and the seat, backrest and feet corresponding to the adjusted equivalent comfort contours generated from the median individual k and n values. (Values calculated using BS 6841 (1987) [10] asymptotic frequency weightings extrapolated below 0.5 Hz without band-pass filtering). Components of motion at the seat, backrest and feet shown together with the rootsums-of-squares of the weighted values in each axis and seating condition.
For fore-and-aft oscillation on a seat with a backrest, the r.s.s. summation of the weighted components produced values which declined from close to the predicted value (0.45 m s 2 rms) as the frequency increased from 0.2 to 1.6 Hz. This implies that the discomfort caused by the higher frequencies was underestimated by the weightings. The principal contribution to the r.s.s was the fore-and-aft oscillation at the seat, with a high contribution from the backrest and a negligible contribution from the oscillation of the feet. For pitch oscillation with no backrest, the seat fore-and-aft acceleration is the dominant term in the r.s.s summation at frequencies less than 0.5 Hz. At greater frequencies, the summation is dominated by the seat pitch component. The root-sum-of-squares of all components are variable but, overall, decrease with increasing frequency. The level at 0.2 to 0.315 Hz is only slightly below the reference condition (0.45 m s 2 rms) but at greater frequencies the reduction indicates that discomfort was underestimated by the standard method. For pitch oscillation with a backrest, the seat fore-and-aft acceleration remains the dominant term between 0.2 and 0.63 Hz. According to the standard method, the response at frequencies greater than 1.0 Hz is dominated by the backrest tangential term. The r.s.s. of the weighted components is at approximately the correct level (0.45 m s 2 rms) at the three lowest frequencies, but declines indicating underestimation of discomfort by the standard method. The standards were evolved from studies with vibration in individual axes, so if one axis dominates they should provide a good prediction. However, this assumes that the weightings for different axes are given the appropriate relative weight and that relative motion between axes, and the phase of any relative motion, does not influence discomfort. In practice the relative motion between inputs (e.g. between the seat and the feet) can have a large influence on discomfort [13,14]. In this case, even though the seat and feet moved in phase with one another, relative motion may still be an important variable since, particularly at the lowest frequencies, there was opportunity for voluntary control over the movement of the upper body, and the relative motion between the seat and upper torso may have been important factors in producing discomfort. 4.7. Implications of the findings ARTICLE IN PRESS I.H. Wyllie, M.J. Griffin / Journal of Sound and Vibration 324 (2009) 453–467 465 Acceleration weighting 2.0 1.0 0.5 0.2 0.1 0.1 1.0 Frequency (Hz) 10 Fig. 7. Asymptotic frequency weightings according to BS 6841 (1987) [10] extrapolated below 0.5 Hz without band-pass filtering: ——— Wb; ?? Wc; ––––Wd; – – – – We. The finding that fore-and-aft acceleration in the plane of the seat is not a reliable predictor of discomfort has implications for the prediction of discomfort in vehicles and vessels. Where there is significant oscillation at
Page 1 and 2: Journal of Sound and Vibration 324
Page 3 and 4: condition (with or without a backre
Page 5 and 6: 2.5. Procedure Prior to participati
Page 7 and 8: Correction factors calculated from
Page 9 and 10: 4. Discussion incidence of dominant
Page 11: Acceleration (rad.s -2 r.m.s.) 10 1
Page 15: ARTICLE IN PRESS I.H. Wyllie, M.J.

Discomfort from sinusoidal oscillation in the pitch and fore-and-aft ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?